Scribd
Upload
64 views87 pages

Chapter 15

aa

Uploaded by

Nevena Vasiljevic
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
64 views87 pages

Chapter 15

aa

Uploaded by

Nevena Vasiljevic
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
(a erent Pate & Sees Capital Structure and the Cost of Capital: Rene as aoa Theory and Evidence of claims on the firm’s cash flows. Debt holders have contracts (bonds) that promise to pay them fixed schedules of interest and principal in the future in exchange for their cash now. Equity holders provide retained earnings or buy rights offerings (internal equity provided by existing shareholders) or purchasenew stares (external equity provided by new shareholders). They do so in return for claims on the residual earnings ofthe firm inthe future. Also, shareholders retain control of the investment decision, whereas bondholdeys have no direct control except for various types of indenture provisions in the bond that may constrain the decision making of sharcholders, Inaddition to these two basic categories of claimants, there are others suchas holders of convertible debentures, leases, prefered stock, nonvoting stock, and warrants, Each investor category is confronted with a different type of risk, and therefore each requires a different expected rate of return in order to provide funds to the firm, The required rate of return is the opportunity cost to the investor of investing scarce resources elsewhere in opportunities with eguivatent risk. AS we shal see, the fact that shareholders are the ones who decide whether to accept or reject new projects is critical to understanding the cost of capital, They will accept only those prajects that increase their expected utility of wealth, Each project must earn, on a ciske adjusted basis, enough net cash flow to pay investors (bondholders and shareholders) their expected rates of return, to pay the principal amount that they originally provided, and to have something left over that will increase the wealth of existing shareholders, The cost of capital is the minimum risk-adjusted rate of return that a project must earn io order to be acceptable to shareholders, ‘Theinvestment decision cannotbe made without knowledge of the cost of capital, Consequently, ‘many textbooks introduce the concept of the cost of capital before they discuss investment deci- sions. It probably does not matter which topic comes first. Both topics are important and they are interrelated. Figure 15.1 shows the investment decision asthe intersection of the demand and sup- ply of investment capital, All projets are assumed to have equivalent risk. Also, fund sources have F UNDS FOR INVESTMENT ae provided to the firm by investors who hold various types 587 38 Chapter 15: Capital Structure and te Cost of Capital Theory and Evidence Figure 15.1 Demand and. % supply of investent for projects of equal risk BR) ———— Marginal cost ' ofcoptal Investment dollars Marginal efiiency of investment equal risk (in other words, in Fig. 15.1 we make no distinction between equity and debt). Chap- ter 2 discussed the ranking of projects assuming that the appropriate cost of capital was known, The schedule of projects with their rates of rotumn is sometimes called the marginal efciency of investment schedule and is known as the demand curve in Fig, 15.1. The supply of capital, rep- resented as the marginal cost of capital curve, is assumed to be infinitely elastic. Implictly, the projects are assumed to have equal risk. Therefore the firm faces an infinite supply of capital atthe rate E(R,) because itis wssumed thatthe projects itoffers are only a small portion of all investment in the economy. They affect neither the total risk of the economy nor the total supply of capita, The optimal investment for the firm is J, and the marginally acceptable project must eam atleast E(R;). All acceptable projects, of course, earn more than the marginal cost of capital. Figure 15.1 isa oversimplified explanation ofthe relationship between the cost of capital and tie amount of investment, However, it demonstrates the interrelatedness ofthe two concepts. Fora given schedule of investments arise in the cost of capita will result in ess investment. This chapter shows how the Firm’s mix of debt and equity financing affects the cost of capital, explains how the cost of capital is related to shareholders’ wealth, and shows how to extend the cost of capital concept 0 the situation where projects do not all have the same risk. If he cost of capital ean be controlled via some judicious mixture of debt and equity financing, then the financing decision ‘can maximize the value ofthe firm Whetheror not an optimal capital structure exists is one ofthe mostimportantissues in corporate finance—and one of the most complex. This chapter covers the effect of tax-deductible debt on the value of the firm, First in a world with only comporate taxes. then by adding personal taxes as well. Next te effect of business disruption and bankruptcy costs is introduced, and we extend the basic Modiglicni-Miller model using the work of Leland. ‘The result is an equilibrium theory of capital structure, The chapter also cavers nonequilibrium theories that include the pecking order theory, signaling, and the effect of forgoing profitable investments. There is also a discussion of the effect of risky debt, warrants, convertible bonds, and callable bonds, Our discussion of optimal capital structure continues by asking two related questions—iiow ean we explain optimal capital structure, if it exists, within an industry, and how can we explain the cross-sectional regularities among industries? Toward this end we survey the empirical evidence. Bankruptcy costs option pricing effects, agency costs and the signaling theory are all discussed A. The Value of the Firm Given Coyporate Taxes Only 559 along with empirical evidence bearing on their validity, Also, the optimal maturity structure of Aebt s presented. Corporate financial officers must decide not only on how mac debt to carry but also its duration, Should it be short-term or long ser deb’ A. - “he Value of the Firm Given Corporate Taxes Only 1. The Value of the Levered Firm Modigliani and Miller [1958, 1963] wrote the seminal papers on the cost of capital, corporate valuation, and capital structure. They assumed either explicitly or implicitly the following: 1. Capital markets ate frictionless, 2. Individuals can borrow and lend at the eisk-free rate 3. There are no costs to bankruptey or to business disruption. 4, Firms issue only two types of claims: risk-free debt and (risky) equity. 5. All firms are assumed to be in the same risk class (operating risk). 6, Corporate taxes are the only form of government levy (ie, there are ao wealth taxes on corporations atid tio personal taxes) 7. All cash flow streams are perpetuities (i... a0 growth. 8. Cosporate insiders and outsiders have the same information (i.e. no signaling opportunities}. 9, Managers always maximize shareholders’ wealth (ie, a agency cost). 10. Operating cash flows are completely unaffected by changes in capital structure I goes without saying that many of these assumptions are unrealistic, but later we can show that relaxing many of them does not really change the majar conclusions of the model of firm behavior that Modigliani and Miller provide. Relaxing the assumption that corporate debt is risk- free will not change the results (see Section D). However, the assumptions of no bankruptcy costs (relaxed in Section E) and no personal taxes (relaxed in Section B of this chapter) ae critical because they change the implications of the model. The eighth and ninth assumptions rule out signaling behavior (because insiders and outsiders have the same information) and agency costs because managers never seek to maximize their own wealth). And the tenth assumption is crucial because the operating cash fiows are not actually independent of capital structure—with the result that investment and financing decisions shoul! be chought of as codeterminant, These issues are discussed in detail tater on “The fifth assumption requires greater clarification, What is meant when we say that all firms hhave the same risk class? The implication is that the expected risky future net operating cash flows, where CF = the risky net cash flow ftom operations (cash flow before interest and taxes), A= aconstant scale factor. s60 CChapter 15; Capital Structure and the Cost of Capital: Theory and Evidence ‘This implies that the expected future cash Rows ftom the two firms (or projects) are perfectly comelated. If, instead of focusing on the level of cash flow, we focus on the returns, the pesfect correlation becomes obvious because the returns are identical, as shown below: and because CF, = 2CF ,, we have ‘Therefore if two streams of cash low differ by, at most, a scale factor, they will have the same distribution of returns, the same risk, and will require the same expected retnrn Suppose the assets of a firm return the same distribution of net operating cash flows each time period for an infinite numberof time periods. This isa no-growth situation because the average cash flow does not change over time, We can value this after-tax stream of cash flows by discounting its expected value atthe appropriate risk-adjusted rate. The value ofthe unlevered firm (ie, firm with no debt) will be EURCR) é (5. where Vip =the preset value ofan unlevered firm (i.e. all equity), E(ECE) = the perpetual fece cash flow after taxes (to be explained in detail below), p= the discount rate for an allequty fmm of equivalent ssk ‘This isthe value of an unlevered firm because it represents the discounted value of a perpetual, rnongrowing stream of free cash flows after taxes that would accrue to shareholders if the fim had ‘no debt. To clarify this point, let us look at the following pro fora statement: Rey Revenues sible eusts of uperations Fixed cash costs (. Noncash charges (€ g, depreciation and deferred taxes) administrative costs and real estate taxes) Earrings before interest and taxes Interest on cet interest cate times principal, D) Earnings before taxes Taxes = 1,(EB7), where r, isthe corporate tax rate MW Netincome A, The Value of the Fira Given Corporate Taxes Only 561 It is extremely important to distinguish between cash flows and the accounting definition of profit, After-tax cash flows from operations may also be calculated as EBIT (earnings before interest and faxes) less cash taxes on EBIT: EBIT — x, €8TT Rewriting this using the fact that EBIT = Rey — VC — FCC — dep, we have (Rev ~ VC - FCC ~ dep) - 1). This is operating income after taxes, but it is mot yet @ cash flow definition because a portion of ‘otal ixed costs are noncash expenses such as depreciation and deferred taxes. Total fixed costs ae partitioned in two parts: FCC is the cash fixed costs, and dep is the noncash fixed costs To conver aiter-tax operating income into cash flows, we must add back depreciation and other noncash expenses. Doing this, we have (Rev — VE — FCC — dep)(t — +,) + dep. Finally, by assumption, we know thal the firm has no growth; that is, all cash flows are perpetuites. This implies that depreciation each year must be replaced by investment in order to keep the same amount of capital in place. Therefore dep = /, and the after-tax free cash flow available for payment to creditors and shareholders is ECF = ier ~ VO— FCC — dep)(1— 1,) + dep - 1, FCF = (Rov — 00 - FCC dep) 1,) since dep =1. The interesting result is that when all eash flows are assumed to be perpetuities, free cash flow (FCA) isthe same thing as net operating income after taxes (ie, the cash flow that the firm woud hhave available if it had no debt at ally. This is shown below: EBITO — 1,) = FCF = (Rev — Fe — FCC - dep — ‘Note also that this approach to cash flows is exactly the same as that used to define cash flows for budgeting purposes in Chapter 2, The reader should keep in mind that in order to determine the value ofthe firm correctly, the definition of cash flows and the definition of the discount rate (ie, the weighted average cost of capital) must be consistent. The material that follows will prove that they are Given perpetual cash flows, Eq. (15.1), the value of the unlevered firm, can be written in either of two equivalent ways: EFC) Vy = ECEBITY = 1) e ° (15.2) From this point forward we shall use the net operating income definition of cash flows in order to be consistent with the language originally employed by Modigliani and Miller, 'The present value of any constant perpetual stream of cash lows is simply the cash ow divided bythe discount rate, See Appendix A at he end ofthe book, Eq (4.5). 562 Chapter 1s: Capital Structure ard the Cost of Capital: Theory and Evidence [Next assume thatthe fin issues debt. The after-tax cash flows must be split up between debt holders and shareholders. Shareholders receive NI + dep ~ I, net cash flows after interest, taxes, and replacement investment; bondholders receive interest on debt, kD. Mathematically this is ‘equivalent co total cash flow available for payment to the private sector: N+ dep — 1 + yD = (Rev — VC - FCC ~ dep — kgDL= 1) + gD. Given that dep =, for a nongrowing fim we can reamrange terms to obtain Ai + kyD = ev — 00 ~ FCC — dep)! — 1.) + ky Dr, (15.3) The first part ofthis stream, EBIT({ — 1), isexacly the same as the cashflows forthe unlevered firm, the numerator of Eg. (15.1), with exactly the same risk. Therefore, recalling that this is a Perpetual stream, we can discount it at the rate appropriate for an undevered firm, p. The second Partof the stream, k, Dr. is assumed to be risk ree, Therefore we shal discount itat the before-tax cost of risk-free debt, ky, Consequenily, the value ofthe levered firm is the sum of the discounted value of the two types of cash flow that it provides: ye — LASBIT I= to wee 1s p js ‘Note thatky D isthe perpetual stream of risk-free payments tobondhelders and that is the current before-ax market-required rate of return for the risk-free stream. Therefore, since the stream is perpetual, the market value of the bonds, B, is px ae (155) ty Now we can rewrite Eq, (15.3) as Vy + teB (15.5) ‘The valve of the levered fitm, Vp is equal tothe valve of an unievered firm, Vi, plus the present value ofthe tx shield provided by debt, r,B. Later on we shall refer tothe “extra” value created by the interest tax shield on debt as the gein from leverage, This is perhaps one of the most important results in the theory of corporation finance obtained in de fast 50 years, It says tha in the absence of any market imperfections including corporate taxes (ie, if 1; =0), the value of the firm is ‘completely independent of the type of Financing used for its projects. Without taxes, we have WeWy. if 1=0. (1550) Equation (15 5) is known as Modigliani-Miller Proposition 1. “The matket valve of any firm is independent ofits capital structure and is given by capitalizing its expected return atthe rate p approptiate to its risk class.”? In other words, the method of financing is irrelevant, Modigliani and Millet went on to suppor their position by using one ofthe very first arbitrage pricing arguments 2-The government receives all cash lows not included in Eq (15.2): thats, he goventment receives taxes (as0 a risky cash ow), Miglin and Miller (1958, 268], ‘A. The Value of the Firm Given Corporate Taxes Orly 563, Table 15.1 Proposition { Arbitrage Example Company A Company 3 EBT 10.000 10,000 “ky D » 1.500, MD 10.000 8.500 4 10% Ne 5 1.000 man B 0 30,000 V=R+S 100,000 107,272 wacc 10% 93% BIS Oe RIG in finance theory. Many say that the arbitrage-free equilibrium was the best of their contributions and the primary reason they deserved the Nobel prize in economics. ‘One of Professor Miller's favorite jokes was a story about the famous baseball player for the New York Yankees, the catcher You Berra It seems that aftera close game he retired with friends to local Ralian restaurant where he ordered an entire pizza for himself. When the waiter asked if Yogi would lke the pizza cut ito six slices or eight, the famous humorist replied, “L woul! fike eight slices please. Lam very hungry.” OF course, the price ofthe pizza was unchanged by how it .was sliced. So 100, argued Modigliani and Miller, the value of firm is independent (aside from tax considerations) of how its liabilities— debt and equity—are partitioned, Consider the income statements of the two firms given in Table 15.1, Both companies have ‘exactly the same perpetual cash flows from operations, ZBI, but company A hes nv debt, whereas company B has $30,000 of debt paying 5% interest. The example reflects greater risk in holding the levered equity of company B because the cost of equity, &, = I'S. for Bis greater than that of company A. ‘The exannple has been constructed so that company B has 2 greater market value than Aaand hence a lower weighted average cost of capital, WACC = EBIT/V. The difference in values isa violation of Proposition J, However the difference will not persist becouse if we already own stock in B, we can earn a profit with no extra risk by borrowing (at 5%) and buying company A. In effect, we create homemade leverage in the following ways 1. We sell stock in B (if we own 1%, then we sel) $772.72), 2. We botrow an amount equivalent to 1% of the debt in B, that is, $300 at 5% interest. 3. We buy 1% of the shares in A. Before arbitrage we held 1% of the equity of B and eamed 11% on it, that i, .11(8772.72) = 85,00. After arbitrage we hold the following postion: {195 of A's equity and eam 10%, that i, 10 ($2,000.00) seon09 pay interest on $30) of debt, that is,.05($300) 15.00 85.00 ‘This gives the same income as our levered postion in company B, but the amountof money we have available is $772.72 (from selling shares in B) plus $300 {from borrowing). So far, in the above 564 Chapter g: Capital Structure and the Cost of Capita: Theory and Evidence calculation, we have used only $1,000.00 to buy shares of A. Therefore we can invest another $72.72 in shares of A and ear 10%. This brings our total income up to $85 + $7.27 = $92.27, and we ovin $772.72 of net worth of equity in A (the bank “owns” $300). Therefore our return 11.94% (ie., $92.27/8772.72). Furthermore, our personal leverage isthe $300 in debe divided by the equity in A, $172.72. This is exactly the same leverage and therefore the same risk as we started with when we had an equity investment in B, ‘The upshot of the foregoing arbitrage argument is that we can use homemade leverage to invest in A. We earn a higher rate of return on equity without changing our risk at all. Consequently, we will undertake the arbitrage operation by selling shares in B, borrowing, and buying shares in A. We will continue to do so until the market values ofthe two firms are identical. Therefore Modigliani-Miller Proposition Iisa simple arbitrage argument. [na world without taxes the market values ofthe levered and unlevered firms must be identical However, as shown by Eq, (15.5), when the government “subsidizes” interest payments to providers of debt capital by allowing the corporation to deduct interest payments on debt as an expense, the market value of the corporation can increase as it takes on more and more (risk-free) debt. Ideally (given the assumptions of the model} the firm should take on 100% debe.’ 2. The Weighted Average Cost of Capital Next, we can determine the cost of capital by using the fact that shareholders will require the rate cof return on new projects to be greater than the opportunity cost ofthe funds supplied by them and bondholders. This concition is equivalent to requizing that original shareholders’ wealth increase. From Eq, (15.3) we see that the change in the value of the levered firm, AV, with respect to a new investment, AJ, is* AV, _ (= to) SEER Alp ar snl “al 5s) Ifwe take the new project, the change in the value of the firm, AVj, will also be equal to the change in the value of original shareholders’ wealth, 4, plus the new equity required forthe project, AS", plus the change in the value of bonds outstanding, 4 B°, plus new bonds issued, A.B": AV, = AS? 4 AS" + ABP 4 AB" (15.7a) Alternatively, the changes with respect to the new investment are ay, _ as’ as" Al ar” At oe Al oe Al (15.70) Because the old bondholders hold a contract that promised fixed payments of interest and principal, because the new project is assumed t0 be no riskier than those already outstanding cand especially because both old and new debt are assumed to be risk free, the change in the value “We shll se ater in vhs chaper that this results mosiid when we considera word with both corporate and personal taxes or on where henfrupey cosis are nontvil. Also, the Itemal Revenue Service wil disallow te tax deductbily ‘ofimerest charges onde if ints judgment, the firm s using excessive debt financing a a ax shied. Nove that ¢ and p do not change with 41. The cost of equity for an allequity firm does not change because new projects, ae assumed to ave the same risks the oid one. A. The Value ofthe Firm Given Corporate Taxes Only 565 ‘of outstanding debt is zero (8° = 0). Furthermore, the new project must be financed with either new debt. new eauity, or both, This implies that® STS AS" + AB" (15.8 Using this fact, Bq. (15.76) cam be veveritten as vy, AS? ASE ABP ase Oe eee (159) ar” Ar al ar For a project to be acceptable to original shareholders, it must increase their weal. Therefore they will cequite that — —i —1>0, (15.10) at al ‘which is equivalent to the requirement that AV,/AJ > 1. Note that the requirement that the change in original shareholders’ wealth be positive (ie., AS°/A/ > 0) isa behavioral assumption imposed by Modigliani and Miller. They were assuming (1) that managers always do exactly what sharehokiers wish and (2) that managers and shareholders always bave the same information. The Jpchavioral assumptions of E (15.10) are essential for what follows ‘When the assumptions of inequality (15.10) are imposed on Ea. (15.7) wear able to determine the cost of capital’ or, by rearranging terms, we have (sp (= 19S€(EBITY ( at , ‘The left-hand side of Eq, (15.11) is the after-tax change in net operating cash flows brought about by the new investinent, that is, the after-tax retum on the project® The right-hand side is the: ‘opportunity cost of capital applicable to the project. As long as the anticipated rate of return om investment is greater than the cost of capital, current shareholders’ wealth will increase. Note that if the corporate tax rate is zero. the cost of capital is independent of capital structure {the ratio of debt to total assets), This result is consistent with Eg (15.5a}, whieh says thatthe value of the firm is independent of capital structure. On the other hand, if corporate taxes are paid, the cost of capital declines steadily asthe proportion of new investment financed with debt increases The value of the levered firm reaches a maximurn when there is 100% debt financing (so long as all the debt is risk free), Note that Ba, (158) does noc require ne issues of debtor equity tobe postive. I is conceivable, for example, that the fim might iss $4,000 in stock for S000 projec and repurchase $309 in debt Note that (AB = AB") because AB" is assumed tobe 20. Chapter 2, the investment decision, stressed ihe point that the correct cash ows for capital budzeding purposes Were always dened as net cashflows from operations afer taxes. Equation (15.11) reiterates this point and show that itis {he only defisition of cash lows that is consistent with the opportunity cost cf capital ar he le. The auriraor ot the Jet-hand side, namely, ECEBIT)(A ~ 1, it the altertascash ows om operations thatthe Firm would have if it had no ooh 366 Chapter 15: Capital Structure and the Cost of Capital: Theory and Evidence 3. Two Definitions of Market Value Weights Equation (15.11) defines what has often been called the weighted average cost of capital, WACC, forthe firm: / wace=p (15,22) (15.12) Amoften-debated question isthe correct interpretation of A 8/A . Modigliani and Miller (1963, 441) imerpret it by saying: If B*/V* denotes the ficm’s long run “target” debt ratio. . then the firm can assiime, to a fist approximation atleast, that for any particular investment dB/di = B*/V*. ‘Two questions arise in the interpretation of the leverage ratio, 4.B/ AY. First, is the leverage ati marginal or average? Modigliani and Miller, in the above quote, set the marginal ratio equal to the average by assuming the firm sets 2 long-run target ratio, which is constant. Even if this is the cas, we sill must considera second issue, namely: Is the ratio to be measured as book value leverage, replacement value leverage, or reproduction value leverage’ Ihe last two definitions, &s wwe shall see, are both market vale At least one ofthese three measures, book value leverage, can be ruled out immediately as being meaningless. In particular, there is no relationship whatsoever between book value concepts (¢g., retained earnings) and the economic value of equity The remaining two interpretations, replacement and reproduction value, make sense because they ae both market value defiitions. By eplacement value, We mean the economic costof putting a project in place, For capita projacts a large part of this cost is usually the cost of purchasing plant, equipment, and working capital In the Modigliani-Miller formulation, replacement cost is the market value of the investment in the project under consideration. AJ. Itis the denominator ‘on both sides ofthe cost of capital inequality (15.11). On the other hand, reproduction value, & Y. js the total present yalue of the stream of goods and services expected from the project. The two concepls can be compared by noting thatthe difference between them is the NPV of he project, that is. NPV=AV-AT For a marginal project, where VPV = 0, replacement cost and reproduction value are equal Haley and Schall (1973, 306-311] introduce an alternative cost of capital definition where the “target” leverage isthe ratio of debt to reproduction value: mec =o {1-122 wace=a(1 At) (15.13) Wf the firm uses a reproduction value concept for its “target” leverage. it will seek to maintain a constant ratio of the market value of debt to the market value of the firm. With the foregoing as background, we can now reconcile the appecent conic in the measure- ment of leverage applicable co the determination of the relevant cost of capital fora new investment project. Modigliani and Miller define the target L* asthe average, in the long run, of te debt(o- value ratio or B*/V*. Then regardless of how a particular investment is financed, the relevant leverage ratio is d8/aV. For example, a particular investment may be financed by debt. But the cost of that particular increment of deb isnot the relevant cost of capital for that investment, The debt would require an equity base, How mach equity? This is answered by the long-run target ‘A. The Value of the Fim Given Corporate Taves Only 567, B*/V*. So procedurally, we start with the actual amount of investment increment for the patic- ular investment, df. The L* ratio then defines the amount of dB assigned to the investment. Ifthe NPV from the investment is postive, then dV will be greater than d7. Hence the debt capzcity of the firm will have been increased by more than dB. However, the relevant leverage for estimating the WACC will still be dB/dV, which wil] be equal to B*/V*. We emphasize that the latter is a policy target decision by the fim based on relevant financial economic considerations, The dV is, an amount assigned to the analysis to be consistent with L”. The issue is whether 10 use dB/dV or dB/di as the weight in the cost of capital formula, The following exaraple highlights the difference between the two approaches. Suppose @ firm can undertake a new project that costs $1,000 and has expected cash flows with a present value of $9,000 when discounted at the cost of equity for an alleguity project of equivalent risk. Mf the ratio ofthe firm’s target debi to value is 50% and if its tax rae is 40%, how much debt should it undertake? If it uses replacement value leverage, then dB/d = .$ and dB = $5(0: that is, half of the $1,000 investment is financed with debt. Using Eg. (15.5) the value ofthe levered fitm is a, = Vy + 1.08 = 9,000 + 41500) = 9,200. ‘The same formula can be used to compute the amount of debt if we use reproduction value leverage, that is, dB/dV = 5, or dV = 24B: d¥, = 9.00 + AUB, 2dB =9,000+ AUB since dV’, =2dB. dB = 5,625. four target is set by using reproduction values then we should issue $5.625 of nev: debt for the $1,000 project, and repurchase $4,625 of equity. The change in the value ofthe firm will be a, = dV" + dB = 9,000 + .4(5625) = 11,250, Clearly, the value of the firm is higher if we use the reproduction value definition of leverage. But as a practical matter, what bank would lend $5.625 on a project that has $1,000 replacement value of assets? Ifthe bank and the firm have homogeneous expectations, this is possible. If they do not, then it is tikely that the firm is more optimistic than the bank about the project. In the case of heterogeneous expectations there is no clear solution to the problem. Hence We favor the original argument of Modigliani and Miler thatthe long-run target debt-to-valve ratio wil be close to dB/dl (ie., use the replacement value definition. 4. The Cost of Equity If Eqs. (15.12) and (15.13) are the weighted average cost of capital, how do we determine the cost of the two components, debt and equity? The cost of debt is the risk-free rate, atleast given the assumptions of this model, (We shall discuss risky debt in Section D.) The cost of equity capital isthe change in the return to equity holders with respect to the change in their investment, 568 CChapter 15: Capital Structure and the Cost of Capital Theory and Evidence AS? + AS", The setum to equity holders is the net cash How after interest and taxes, WE Therefore their rate of return is ANT/(AS* + 45"), To solve for this, we begin with identity (15.2), M+ kyD = EBIT(| - 1.) + &yDre Next we divide by AJ, the new investment, and obtain ANE jA0D)_teOD) 7) ,) SEE. as.14) ar” AL Ar Al Substituting the left-hand side of (15.14) into (15.6), we get 2M, _ AMAT += cgay y/Or AB a al ? al From (15.7), we know that AV, _ 49745" | apt et al ar Consequently, by equating (15.15) and (15.16) we get AY, _ A +AS" | AB AN/AL+(-r)AlKyDI/AI | AB ar al A > “Al ‘Then, multiplying both sides by AJ, we have asty ast apa SNAG =e) thgd) + or AB, ? Suiblracting &B from both sides gives ase past 2 ANTEC t Ady) + 61,48 = pas ? P(AS? FAS") = ANI —(1-r)(D~hy)AB, since (kD) = Ay AB, And finally, _ NI ab a ap (1-9 - 10517) aerax? NP - ae ‘The change in the new equity plus old equity equals the change in the total equity of the firm (AS = AS? + AS*), Therefore the cost of equity, k, = ANI/AS, is writen 4, a8 +(1= fp — by) (15.18) e ne ‘The implication of Eq. (15.18) is that the opportunity cost of capital to shareholders increases linearly with changes in the market value ratio of debt to equity (assuming thar AB/ AS = B/S) If the firm has no debt in its capital structure, the levered cost of equity capital, é,, is equal to the cost of equity for an all-equity firm, p. ‘A, The Value of the Firm Given Corporate Taxes Only 569 152 The cost of capital as a function of the ratio of debt to equity: (a) assuming r, (b) assuring F, > 0. « % B Ka p+(l-1¥p-k)e fa) (b) 5. A Graphical Presentation for the Cost of Capital Figuce 15.2 graphs the cost of capital and its components cs a function of the ratio of debs to equity ‘The weighted average cost of capital is invariant to changes in capital stucture in a world without cconporate taxes; however, with taxes it declines as more and more debt is used in the firm's capital structure. In both cases the cost of equity capita inereases with higher proportions of debt. This, makes sense because increasing financial leverage implies a riskier position for shareholders as their residual claim on the firm becomes more variable. They require a higher cate of return to compensate them forthe extra risk they take, ‘The careful reader will have noticed that in Fig, 15.2 B/S is on the horizontal axis, whereas Eqs. (15.13) and (15.18) are written in terms of AB/AS or AB/A, which are changes in debt with respect to changes in equity or value of the firm. The two are equal only when the firm's average debt-to-equity ratio is the same as its marginal debt-to-equity ratio. This will be true as long as the firm establishes a “target” debt-to-equity ratio equal to B/S and then finances all projects with the identical proportion of debt and equity so that B/S = AB/AS. ‘The usual definition of the weighted average cost of capital is to weight the after-tax cost of debt by the percentage of debt in the firm's capital structure and ad the result tothe cost of equity ‘multiplied by the percentage of equity. The equation is WAC = Ah raha +, S (15.19) +50 "B45 ‘We can see that this is the same as the Modigliani-Miller definition, Eq, (15.12), by substituting (15.18) into (15.19) and assuming tat B/S = B/S. 570 Chapter 1s: Capital Structure and the Cost of Capital: Theory and Evidence a B 5 HACE (= thy 8 — {0-4 (1~ 29 — bod | { hers [r (tip 2S, nee 8 —— PS Bas *SB4+S s B B 8 wi FP) pe 8 eg bes rs) Pays Oba B =p{l-r—=-). Qe of a) s ‘There is no inconsistency between the traditional definition and the M-M definition of the cost of capital (Eqs. (15.12) and (15.19)}. They are identical Pa B. Tre Value of the Firm in a World with Both Personal and Corporate Taxes 1. Assuming All Firms Have Identical Effective Tax Rates Jn the original mode! the gain fiom leverage, G, i the difference between the value of the levered and unlevered firms, which isthe product ofthe corporate tax rate and the market value of debt: GaV,-Vys08 (1520) Miller(1977] modifies this result by introdueing personal as well as corporat taxes into the model. In addition 1o making the model more realistic, the revised approach adds considerable insight into the effect of leverage on value in thereal world, We da not, afterall, observe firms with 100% debt in their capital structure as the original Modigliani-Miller model suggests. Assume for the moment that there are only two types of personal tax rates: the rate on income received from holding shares, t,,, and the rate on income from bonds, tpg. The expected after tax stream of cash flows to shareholders ofan all-equity firm woutd be (EBIT)(I ~ 1,)(0 — tye) By discounting this pespetual stream atthe cost of equity for an all-equity firm, we have the value of the unlevered firm: Vp ELEM = 1 U= ty w (1s21) 9 Alternatively if the firm has both bonds and shares outstanding, the earnings stteam is pari- tioned into iwo pacts. Cash flows to shareholders after corporate and personal taxes are payments to shareholders = (EBIT — 4,B)(4~ .) = ty.) and payments to bondholders, after personal taxes, arc payments to bondholders = ky (I ~ pp). ‘Adding these together and rearranging terms, we have B. The Value of the Firm in a World with Both Personal and Corporate Taxes $72 total cash payments to suppliers of capital = EBIT() — M1 = typ) — RDU — GHA ty) + RgDCE~ tpg). (15.22) ‘The first term on the right-hand side of (15,22) is the same as the stream of cash flows to owners of che unlevered firm, and its expected value can be discounted atthe cost of equity for an allequity firm, The second and third terms are risk free and can be discounted at the risk-free rate, ky. The su of the discounted streams of cash flow is the value of the levered firm: yp = FBIM = 100 ty), Hy [= tpn) = 0-190 : > by 1=1)0- 4 [t= SE), (1523) =the) where B= kgD(1~ t)g)/kyy the market value of debt. Consequently, with the introduction of perstmal taxes, the gain from leverage is the second term in (15.23): (l-1)(-4, eed la (1528) Gai =a Note that when personal tax rates are set equal to zero, the gain from leverage in (15.24) equals the gains from leverage in (15.20), the eacier result. Ths finding also obtains when the personal tax rate om share income equals the rate on bond income. In the United States it is reasonable to assume thatthe effective tax rate on common stock is lower than that on bons The implication is thatthe gein from leverage when personal taxes are considered (15.24) is lower than zB (15.20), Ifthe personal income tax on stocks is less than the tax on income from bonds, then the before- tax sevwrn on bonds has to be high enough, other things being equal, to offer this disadvantage Otherwise the investor would want to hold bonds, While it is true that owners of a levered corporation ate subsidized by the interest deduetibility of deb, this advantage is counterbalanced) by the fact thatthe required interest payments have already been “grossed up” by any differential that bondholders must pay on their interest income, In this way the advantage of debt financing may be lost In fact, whenever the fotiowing condition is met in Eq, (15.24), (= ta) = (1-10 tp) (1525) the advantage of debt vanishes completely Suppose that the personal tax rate on income from common stack is zero. We may justify this by arguing that (1) no one has to realize acapital gain until after death: (2) gains and losses in well- diversified potfalios can offset cach ollea, hereby eliminating the payment of capital gains tes; (3) 80% of dividends received by taxable corporations can be excluded from taxable income; or (4) many types of investment funds pay no taxes at all (nonprofit organizations, pension funds, trust funds, etc,)'° Figure 15.3 portrays the supply and demand for corporate bonds. The rate paid ‘on the debt of tax-free institutions (municipal bonds, for example) is ¥o, Ifall bonds paid only rp, * Tetacrate onsoekis thought of as being lowe than hat on bonds because ofa relatively igh capital gains comaponent of return, and because cpitl gins ae not taxed Ut he security is sol. Capital gins taxes can, therefore, be deemed inden. "© Also, 4s will be shown in Chapter 7, is possible ose upto $10.00 in dividend income fem aes. 572 Chapter 1g: Capital Structure and the Cost of Capita: Theory and Evidence Figure 15.3 Aggregate % supply and demand for | corporate bonds (before tax rates) Supply i 1 " ! I i —<— — : ~ Dollar amount of all bonds no one would hold them, with the exception of tax-free institutions that are not affected by the tax advantage of holding debi wher z,g > Tp. An individual with a marginal tax rate on income ‘rom bonds equal tor} will not hold conporate bons until they pay rp/(— fg), that i, until their retumn is “grossed up.” Since the personal income tax is progressive, the interest rate that is demanded has to keep rising to attract investors in higher and higher tax brackets."! The supply of ‘corporate bonds is periectly elastic, and bonds must pay a rate of ro/( ~ ¢,) in equilibrium. ‘To see that this is true, let us recall that the personal tax rate on stock is assumed (o be zero (ys = 0) and rewrite the gain from leverage: G=(I- B. 15.26) (-253) al If the rate of return on bonds supplied by corporations is r, =ro/(1— 1), then the gain from leverage, in Eq, (15.26), will be zero, The supply rate of return equals the demand rate of return in equilibrium: Consequently, (= =U tpah and the gain from leverage in (15.26) will equal zero. If the supply rate of return is less than ro/(1— 1), then the gain from leverage will be postive, and all corporations will try to have a capital structure containing 100% debt. They will rush out to issue new debt, On the other hand, if the supply rate of return is greater than r/(I — r,), the gain from leverage will be negative and fisms will take action to repay outstanding debt. Thus we see that, in equilibrium, taxable debt must be supplied tothe point where the before-tax cost of corporate debt must equal the rate that ‘would be paid by tax-free institutions grossed up by the corporate tax rate. Keep in mind that the tx rate on income from stock is assumed to be zero. Therefore the higher an individuals tax bracket becomes the higher the before-tax rae on onds must be inorder for he afie-tax rate on bonds to equal the rate ‘of return on sto (after ajstng fo isk), B, The Value of the Firm in a World with Both Persoral and Corporate Taxes $73, Miller's argument has important implications for capital structure, First, the gain to leverage may be much smaller than previously thought. Consequently, optimal capital structure may be explained hy a trade-off between a small gain to leverage and relatively small costs suck as expected bankruptcy costs. This trade-off will be discussed at greater length in the book. Second, the observed market equilibrium interest rate is seen to be a before-tax rate that is “grossed up” so that most or all of the interest tax shield is Jost, Finally, Miller's theory implies there is an equilibrium amount of aggregate debt outstanding in the economy that is determined by relative corporete and petsonal tax rates 2, Assuming That Firms Have Different Marginal Effective Tax Rates ‘DeAngelo and Masulis [980] extend Miller's work by analyzing the effect of tax shields other than interest payments on debt (e.g., noncash charges such as depreciation, oil depietion allowances, and investment tax credits). They ae able to demonstrate the existence of an optimal (nonzero) ccomporete use of debt while still maintaining the assumption of zero hankruptcy (and ze10 agency) costs ‘Their original arguments illustrated in Fig. 15.4. The corporate debt supply curve is downward sloping to reflect the fact that the expected marginal effective tax rae, r/, differs across corporate suppliers of debt. Investors with personal tax rates lower than the marginal individual ear a consumer surplus because they receive higher after-tax retumns. Corporations with higher tax rates, than the marginal frm receive a positive gain to leverage, a producer's surplus. in equilibrium because they pay what is for them a low pretax debt rate Itis reasonable to expect depreciation expenses and investment tax credits to serve as tax shied substitutes for interest expenses. The DeAngelo and Masulis model predicts that firms will select a level of debt that is negatively related to the Level of avzilable tax shield substitutes such as depreciation, depletion, and investment tax credits. Also, as more and more debt is utilized, the probability of winding up with zero or negative earnings will increase, thereby causing the inietest tax shield 10 decline in expected value. They further show that if there are positive bankruptey costs, there will be an optimum trade-off between the marginal expected benefit of interest tax shields and the marginal expected cost of bankruptcy. This issue will be further discussed in the next chapter, Figure 15.4 Agregue % debt equilibsium with heterogeneous corporate and personal tax rates. “Producer surplus” meee bel Pa ; , g ( 1 “| of all bonds 574 Chapter 15: Capital Structure and the Cost of Capital: Theory and Evidence Introducing Risk—A Synthesis of the Modigliani-Miller Model and CAPM The CAPM discussed in Chapter 6 provides anatural theory for ie pricing of isk. When combined with the cost of capital definitions derived by Modigliani and Miller (1958, 1963}, it provides a ‘unified approach to the cost of capital. The work thal we shall describe was first published by Hamada [1969] and synthesized by Rubinstein (1973). The CAPM may be written as F(R) = Ry + (ER) — RylB. sz where E(R;) =the expected rate of retum on asset j, Ry =the (constant) riskefree rate, E(R,) = the expected rate of retum on the marke: portfolio By =COV(R,, R,)/VAR(R,) Recall that all securities fall exactly on the security market line, which is illustrated in Fig. 15.5, ‘We can use this fact io discuss the implications forthe cost of debs, the cost of equity, the weighted average cost of capital, and for capital budgeting when projects have different risk Figute 15.5 illustrates the difference between the original Modigliari-Miller cost of capital and the CAPM, Modigliani and Miller assumed that all projects within the firm had the same business ‘or operating risk (mathematically, they assumed that CF, ~ CF). This was expedient because in 1958, when the paper was written, there was no accepted theory that allowed adjustments for differences in systematic risk. Consequently, the Modigliani-Miller theory is represented by the horizontal fine in Fig, 15.5, The WACC forthe firm (implicitly) does not change as a function of systematic risk. This assumption, of course, must be modified because firms and projects differ in isk. Figure 155 The security ER) soatket line. ; Security market line EReyy) WACC Cir) &, FUR} . Introducing Risk—~A Synthesis of the Modigliani-Miller Model and CAPM 575, Table 15.2, Comparison of M-M and CAPM Cost of Capital Equations ‘Type of Capital CAPM Definition M-M Definition Debt by= Ry +1E (Ry) — Rylb ky= Ry B= Unlevered equity 7 + LELR,) = Ry By pap = Ry + [ER = Ry, k=ptip WACC for the firm WACC = (0b 1.) 985 +h wace=(1- Levered equity 1. The Cost of Capital and Systematic Risk ‘Table 15.2 shows expressions for the cost of debt, k,, unlevered equity, p, levered equity, k, and the weighted average cost of capital, WAC, in both the Modigliani- Miller (M-M) and capital asset pricing model frameworks, t has already been demonstrated in the proof following Eq. (15.19), that the traditional and M-M definitions of the weighted average cost of capital (the last line in ‘Table 15.2) are identical, Modigliani and Miller assumed, for convenience, that corporate debt is risk free; tha is, its price is insensitive to changes in interest rates and either that it has no default risk oF thot default risk is completely diversifiable (f, = 0). We shall temporarily maintain the assumption that ky = R then rela ita Fite later in the chapter The M-M definition of the cost of equity for the unlevered firm was tautological (ie., p = p) because the concept of systematic risk had not been developed in 1958. We now know that it depends on the systematic risk of the firm's after-tax operating cash flows, By, Unfortunately for empirical work, the unlevered beta is not directly observable. We can, however, easily estimate the levered equity beta, By. (This has also been referred to as B, elsewhere.) If there isa definable relationship between the two betas, there are many practical implications (as we shall demonstrate with a simple numerical example in the next section), To derive the relationship between the levered and unlevered betas, begin by equating the M-M and CAPM definitions of the cost of levered equity (tine 3 in Table 15.2): Ry + [E(Rq) — Ry] Bp = 9 +49 — kV -w8 Next, use the simplifying assemption that ky = Ry to write Ry + [ELRn) ~ Ry] Bu = 0 + (0 ~ Ry oe ‘Then substitute into the right-hand side the CAPM definition ofthe cost of unlevered equity, p Ry +[BlRq) ~ Re] Be = Ry + [BRy)~ Rp] By + [Fy + [Eley - Ry] fo -h]-0e 576 Chapters; Capital Structure and the Cost of Capital: Theory and Evidence By canceling terms and reatranging the equation, we have I, B TE (Ry) ~ RyVBr = [E(Bn) — Ry] [+O -w05 | Be» n=[!+0- wo] a (1528) The implication of Eg. (15.28) is that if we can observe the levered beta by using observed rates of return on equity capital in the stock market, we can estimate the unlevered risk of the firm’s operating cash flows. 2. A Simple Example ‘Theusefulness ofthe theoretical results can be demonstrated by considering the following problem. ‘The United Southern Construction Company currently has a market value capital structure of 20% debt to total assets. The company’s treasurer believes that more debt can be taken on, up to a limit of 35% debs, without losing the firm's ability w borrow at 7%, the prime rate (also assumed w be the risk-free rate) The firm has a marginal tax rate of $0%, The expected return on the market vent year is estimated to he 17%, and the systematic risk of the company’s equity, fis estimated tobe S. + Whats the company’s current weighted average cost of capital? Its current cost of equity? + What will the new weighted average cost of capital be ifthe “target” capital structure is changed 10 35% debt? * Should a project with a 9.25% expected rate of return be accepted if its systematic risk, i, is the same as that of the firm’? ‘To calculate the company’s current cost of equity capital, we can use the CAPM: 4, = Ry + [B(R,) — Ry] By =.074 (17-07) 5=.12. ‘Therefore the weighted average cost of capital is s B+S == 5.012) + 12(8) = 10.3%. WACC = Waray pt ty ‘The weighted average cos! of capital with the new capital structure is shown in Fig. 15.6.!? Note that te east of equity increases with increasing leverage. This simply reflects the Fact that shareholders face mote risk with higher financial leverage and that they require a higher return to compensate them for it, Therefore in order to calculate the new weighted average cost of * Note tha if debe to total esses is 20%, then debt to equity is 25%. Ako, 39% convers to 53.89% in Fig. 156. . Introducing Risk—A Synthesis ofthe Modigliani-Miller Model and CAPM 577 Figure 156 Changes in % the cost of capital as | Kapri -tyo-hye leverage increases. | wacceptt-1,-B- wacc=p(l—1, 52s) 25% 53.85% . capital we have to use the Modigliani-Miller definition to estimate the cost of equity foram all- equity firm: a) B+S, wace p= — re [B/(B + 5)] As long as the firm does not change its business tsk, its unlevered cost of equity capital, p, will not change. Therefore we can use p to estimate the weighted average cost of capital with the new capital structures WACC =. 1144{1 ~ 5(35)] = 9.438%. ‘Therefore, the new project with its 9:25% rate of return will not be acceptable even if the firm increases its ratio of debt to total assets from 20% 10 35%. A common error made in this type of problem is (o forget that the cost of equity capital will increase with higher leverage. Had we estimated the weighted average cost of capital, using 12% for the old cost of equity and 35% debt'as the target capital stricture, we would have obtained 9.03% as the estimated weighted average cost of capital, and we Would have accepted. the project. We can also use Eq. (15.28) to compute the urlevered bera for the firm. Before the capital structure change, the levered beta was fly = 5, Userefone B o.=[)40- | bs S=[1+ (1~5(.25)] By, By = 4444, 58 Chapter 15; Capital Structure and the Cost of Capital: Theory and Evidence Note thatthe unleyered beta is consistent with the fira’s unfevered cost of equity capital. Using the CAPM, we have p= Ry +[B(Rn) — Ry) By = 07+ [17-07] = 14d, Finally, we know that the unfevered beta will not change 2s long as the firm does not change its business tisk, the risk of the portfolio of projects that it holds. Hence an increase in leverage will increase the levered beta, but the unlevered beta stays constant. Therefore we can use Bi, (15.28) to compute the new levered equity beta: B =]led-15) ay B= |1+¢ ot) = [1+ (1 -0.5).5385) 4444 = S641 Hence the increase in leverage raises the levered equity beta from 5 10 5641, and the cost of levered equity increases from 12% to 12.64%. 3. The Cost of Capital for Projects of Differing Risk A more difficult problem is to decide what to do ifthe projects risk is different from that ofthe firm. Suppose the new project would increase the replacement market value of the assets of the frm by 30% and che systematic risk ofthe operating cash flow it provides is estimated to be By, = 1.2 What rae of return must it eam inorder tobe profitable ifthe firm has (a) 20% or (b) 35% debt, its capital structure?” Figure 15.7 shows that the CAPM may be used to find the required rate of return given the beta of the project without leverage, By,» which has been estimated tobe 1.2, This is the beta forthe unlevered project, because the betas defined asthe systemratc risk ofthe operating cash flows. By definition this isthe covariance between the cash flows before leverage and taxes and the matket index; divided by the variance of the market portfolio. The required rate of return on the project, if is an all-equity project, will be computed es ER, Ry +[E(Rp) ~ Ry] Buy = O74 [17 = ON L2= 19%. Next we must “add in” the effect of the firm's leverage. If we recognize that 19% is the required rate if the project were all equity, we can find the required rate with 20% leverage by using the Modigliani-Miller Weighted average cost of capital, Eg. (19.1 B wee =o(1=1522) = 19 = 502) 17.1%. ‘And ifthe leverage is increased to 35%, the equired retnfalls to 5.675%: WACC = .19{1 ~ 5(.35)]= 15.675%, D. The Costof Capital with Risky Debt $7 Figure 15.7 Using the EAR) CAPM to estimate the required rate of return on a project, F(R) = 19% BAR, )= 11% b= Firms seek to find projects that earn more than the project's weighted average cost of capital Suppose tha, forthe sake of argument, the WACC of out firm is 17%. Project B in Fig. 15.7 cams 20%, more than the firm's WACC. whereas project A in Fig. 15:7 eams only 15%. which is less than the firm’s WACC. Does this mean that B should be accepted while A is rejected? Obviously nol, because they have different risk (and possibly different optimal capital structure) than the firm asa whole, Project B is much riskier and must therefore earn a higher rate of return than the firm In fact it must earn more than projects of equivalent risk, Since it falls below the security market line, itshould be rejected. Alternately, project A should be accepted because its anticipated rate of return is higher than the rate that the market requires for projects of equivalent risk. It lies above the security market line in Fig. 15.7 ‘The examples above serve to illustrate the usefulness ofthe risk-adjusted cost of capital for capital budgeting purposes. Each project must be evaluated at a cost of capital that reflects the systematic risk of its operating cash flows as well as the financial leverage appropriate for the project. Estimates of the correct opportunity cost of capital are derived from a thorough understanding of the Modigliani-Miller cost of capital and the CAPM. D. Ne Cost of Capital with Risky Debt So far it has been convenient to assume that corporate debt is risk free. Obviously it is not Consideration of risky debt raises several interesting questions. Firs, if debt is risky, how are the basic Modigliani-Miller propositions affected” We know that riskier debt will require higher res of return. Does this reduce the tax gain from leverage? The answer is given in Section 1 below. The second question is, How can one estimate the required rate of return on ricky debt? This is covered in Section 2, 1. The Effect of Risky Debt in the Absence of Bankruptcy Costs ‘The fundamental theorem set forth by Modigliani and Miller is that, given complete and perfect, capital markets, it does not make any difference how one splits up the stream of operating cash flows. The percentage of debt or equity does not change the total value of the cash stream provided by the productive investments of the firm. Therefore, so long as there are no costs of bankrupley 580 ‘Chapter §: Capital Structure and the Cost of Capital: Theory and Evidence (paid to third parties like trustees and law firms), it should not make any difference whether debt is risk free or risky. The value of the firm should be equal to the value of the discounted cash flows from an investment. A partition that dives these cash flows into risky debt and risky equity has ‘no impact on vaiue. Stiglitz {1969} first proved this result, using a state preference framework, and Rubinstein {1973] provided a proof, using a mean-variance approach, Risky debi, just like any other security, must be priced in equilibrium so that it falls on the security market line. Therefore, if we designate the return on risky debt as Ryy, its expected retum is, ECR) = Ry + [ECR — Ry] By (15.29) where f,; = COV(Ryj, Ry)/o. The retum on the equity of a levered fiem, A, ean be written (for 4 perpetuity) as net income divided by the market value of equity: (Pir - RB 2) 7 (5%) Recall that EBIT is earnings before interest and taxes, iB is the interest on debt, r; is the firm's tax rate, and $” is the market value of the equity in a levered firm, Using the CAPM, we find that the expected return on equity will be!® EG) =Ry + MCOV(K, Rn. as31) ‘The covariance hetween the expected rate of return on equity and the market index is —p {[GBTRopBIO = te) (EBIT ~ Ry BNL =r) COVE, R,) = E ({—- -E ( ee) x [@qy ~ E(Ry))} 1 Beov EBir Ry) aati 8 covery. Ra) (1532) Substituting the result into (15.31) and the combined result into (15.30), we have the following relationship fora levered fim: Ry St + AY ~ F JCOV(EBIT, Ry) = = E(EBITY(L ~ ¢.) - E(Ry) BA ~ ¢). (1533) = 1 )BICOVIRy. Ry By following a similar line of logic for the unlevered firm (where B = 0, and S* = V"), we have RyV" + AML =r )COV(EBIT, R,,) = E(EBIZ}(N ~ 12) (8.34) Substituting (15.34) for E(EBIT)(1 ~ ,) in the right-hand side of (15,33) and using the fact that +B, wehave [ECRy) ~ Ry] fo2. D. The Cost of Capital wth Risky Debt st RySy + MU ~ 1 QCOVIEBIT. Rog) — 2° (1 = %)BICOVEA,;, Ry) = RyVy + °C — JCOVEEBIT, Ry) ~ E(Ry IL JR), = B) MU = 1 )BICOVER,,, Ryd = RyVy IR, +N COVER). Ry NIBU = v iy + teB ‘This is exactly the same Modigliani-Mitier result that we obtained when the firm was assumed to issue only risk-free debt. Therefore the introduction of risky debt cannot, by itself, be used to explain the existence of an optimal capital structure. Later on, we shall see that direct bankruptcy ‘costs such as losses to third parties lawyers or the courts) or business disruption costs (disruption of services to customers or disruption of the supply of skilled labor) are necessary in conjunction with risky debt and taxes in order to explain an optimal capital structure, 2. The Cost of Risky Deb!—Using the Option Pricing Model Even though risky debt without bankruptey costs doesnot alter the basic Modigliani-Millerresults, we are still interested in knowing how the cost of risky debt is affected by changes in capital structure, The simple algebraic approach that follows was provided by Hsia [1981], and it combines the option pricing model (OPM), the capital asset pricing model (CAPM), and the Modighiani- Miller theorems. They are all consistent with one another, ‘To present the issue in its simplest form, assume (1) thatthe firm issues Zero- coupon bonds! that prohibit any capital distsibutions (such as dividend payments) until after the bonds mature T time periods hence, (2) that there are no transactions costs or taxes. so that the value of the firm is unaffected by its capital structure (in other Words, Modigliani-Miller Proposition [is assured to be valid), (3) that there is a kaown vonstochastic risk-free tate of interest, and (4) that there are homogeneous expectations about the stochastic process that describes the value of the firm's assets. Given these assumptions, we can imagine a simple firm that issues only one class of bonds, secured by the assets of the frm. To illustrate the claims of debt and shareholders, let us use put-call parity from Chapter 7. ‘The payoffs from the underlying risky asset the value of the firm, V) plus put writen om it are identical to the payoffs from a default-fee zero-coupon bond plus a call (the value of shareholders” equity in a levered firm, S) on the risky asset. Algebraically this is the same put-call parity relationship that was described in Chapter 7: 4 V+PaBts, or rearranging, V=(B-P)+S. (1535) Equation (15.35) illustrates that the value of the firm can be partitioned into two claims. The low= risk claim is risky debt that is equivalent to default-free debt minus a put option, that is, (2 — P) “Thus, risky comporate debt is the same thing as default-free debt minus 2 put option. The exercise "All acclaimed interest on zero-coupon bards is paid at maturity; hence BT), the curene markt value of Jet with maturity 7, ust be less than its face value, D, assuming a positive risk-free rate of discount. 582 Chapter 15: Cepital Structure and the Cost of Capital: Theory and Evidence Table 153 Stakeholders’ Payoits at Maturity Payoffs at Maturity ‘Stakeholder Positions WWsD WV>D Shareholders’ postion Cal option, S| 0 v-D Bondholders’ position: Defautt-fiee bond, B D D Minus a put option, P -(b-¥) ‘Valve ofthe Firm at moturity v price for the pat isthe face value of debt, D, and the maturity of the put, 7, isthe same as the maturity of the risky debt, The higher-risk claim is shareholders’ equity, which is equivalent to a callon the value of the firm with an exercise price and a maturity 7.. The payotf to shareholders at maturity will be S=MAX\0, V =D), (15.36) Table 15.3 shows both stakeholders’ payoffs at maturity. I the value of the firm is less than the face value of debt, shareholders file forbznkruptcy ard allow the hondholders to keep V < D. Alternately ifthe value ofthe firm is greater than the face value of debt, shareholders will exercise their cil option by paying its exercise price, D, the face value of debt to bondholder, and retain the excess value, V — D. The realization thatthe equity and debt in a firm can be conceptualized as options allows us to use the insights of Chapter 7 on option pricing theory. For example, ifthe equity, S, in a levered firm is analogous to 2 call option, then its value will increase with (1) an increase in the value of the irm’s assets, V, (2) an increase in the variance ofthe value of the firm's assets, (3) an increase in the time to maturity of « given amount of debt with face value, D, and (4) an iacrease in the risk-free rate. The value of levered equity will decrease with a greater amount of debt, D, which is analogous to the exercise price ona cal) option. Next, we wish to show the relationship between the CAPM measure of risk (ie., @) and the ‘option pricing model. First, however, itis useful to show how the CAPM and OPM are related, Merton [1973] has derived a continnous-time version of the CAPM: El) =r¢ + (Eq) —r91B:. 5.37) where E(r,) = the instantaneous expected rate of return on asset i, {i= the instantaneeus systematic risk ofthe ith asset, f; = COV(r, 1 )/WARU in) (ra) = the expected instantaneous rate of return onthe market porfoio, +p the nonstochastic instantaneous annualized rate of return on the risk-free asset. ‘There appears to be no difference between the continuous-time version of the CAPM and the traditional one-period model derived in Chapter 6. However, it is important to prove that the CAPM. D. The Cost of Coptal wth Risky Debt 583 also exists in continuous time because the Black-Scholes OPM requires continuous trading, and the assumptions underlying the two models must be consistent. In order to relate the OPM to the CAPM it is easiest (believe it or not) to begin with the differential equation given in Appendix 7A, at the end of Chapter 7, Eq. (7.2), and to recognize that the call option is now the value ofthe common stock, S, which is written on the value of the levered firm, V. Therefore Eq. (7A.2) may be rewritten 2s as 9S 1 aS dS = dV + dr + = ave it + ovat 1538) at 22av2 y y ‘This equation says that the change in the stock price is related to the change in the value of the firm, dV, movement of te stock price across time, di, and the instantaneous Variance of the firm's value, 02, Dividing by 5, we have, in the limit as dv approaches 2e70, dS _bSdV_ aSdVV i 1529 aes BV SS a We recognize dS/5 as the rate of return on common stock, rs, and dV/V as the rate of return on. the firm's assets, ry; therefore ety (15.40) If the instantaneous systematic risk of common stock, is, and that of the firm’s assets, Ay, are defined as COV tind _ COVEY Fn) SET VAR)! VARUR) wa) then we can use (15.40) and (15.41) to rewrite the instantaneous covariance as (1542) Now write the Black-Scholes OPM where the call option is the equity of the firm: S=VNid) ~e/ DNids), (1543) ‘where S= the market value of equity, V = the market value ofthe fis asses, ry = the risk-free rate T = the time to maturity, D = the face value of debt (book value), ‘N (>) = the cumulative normal probability of the unit normal variate d), InV(D) +ryT 1 d= ———+ + oT, I aif yey, y= dy-ovT, Chapter 1: pital Structure and the Cost of Capital: Theory and Evidence Finally, the partial derivative of the equity value, 5, with respect to the value of the underlying assets is S =Nian where 0< Ma) <1. asa) ‘Substituting this into (15.42), we obiain v B= MidB (15.45) This tells us the relationship between the systematic risk of the equity, fs, and the systematic risk of the ficm’s assets, By. The value of $ is given by the OPM, Bq, (15.43), therefore we have a VN) ~ ¥N(d,) = Deo?™ N(d;) 1 “wap eal Bs ‘By (13.46) We know that D/V < I, that e~’” < 1, that (dy) < N (d)), and hence that fy > By > 0. This shows that the systematic risk of the equity ofa levered firm is greater than the systematic risk of an unlevered firm, 2 result that is consistent with the results found elsewhere in the theory of finance. Note also that the beta of equity af the levered fitm increases monotonically with leverage. ‘The OPM provides insight into the eifect of its parameters on the systematic risk of equity. We ‘may assure that the risks of characteristics ofthe firm's assets, By, are constant over time. Then it can be shown that the partial derivatives of (15.46) have the following signs: Bs co, Miso, Bs cg, Bs co, Beco av ap ny fio? ar ‘Most of these have readily intuitive explanations. The systematic risk of equity falls as the market value of the firm increases, and it rises as the amount of debt issued increases. When the risk-free rate of return increases, the value of the equity option increases and its systematic risk decreases, ‘The fourth partial Gerivative says thet as the variance ofthe value of the firm’s assets increases, the systematic risk of equity decreases. This result follows from the contingent claim nature of equity. ‘The equity holders will prefer more variance to less because they profit from the probability tat the value of the firm will exceed the face value of the debt. Therefore their risk actually decreases asthe variance of the value of the firm’s assets increases '5 Finally, the fifth partial says that the systematic risk of equity declines as the maturity date of the debt becomes longer and longer. From the shareholders” point of view, the best situation would be to never have to repay the face value of the debt. Itis also possible to use Eo. (15.45) to view the cost of equity capital in an OPM framework and to compare it with the Modigliani-Miller resalts ‘Sow that since the value ofthe frm, Vand the debi equity ratio Dj V ae held const, any change in total variance, ‘o?, must be nonsystematc risk. D. The Cost of Capital with Risky Debt 585, Substituting fs from (15.45) into the CAPM, we obtain from Eq, (15.37) an expression fork, the cost of equity capital: v Ry + Ry — RN) By (sary Note that from Eq. (15.45), As = N(dj)(V/S)By- Substituting this into (5.47) yields the familiar CAPM relationship k, = By + (Ry — Ry)s. Furthermore, the CAPM can be rearranged to show that vr Ry Wee which we substitute into (15.47) to obtain v Ry + NCdIRy — Ry) ¢ (3.48) Equation (15.48) shows that the cost of equity is an increasing function of financial leverage. If we assume that debt is risky and assume that bankruptcy costs (ie, losses to third parties ‘other than creditors or shareholders) are zero, then the OPM, the CAPM, and the Modigliani- Miller propositions can be shown to be consistent, The simple algebraic approach given below ‘was proved by Hsia [1981]. First, note that the systematic risk, ig, of risky debt capital in a world without taxes can be written in an explanation similar to Eg, (15.42) as!® aBY =py——. 15.49) p= By we (15.49) We know that in a world without taxes the value of the firm is invariant to changes in its capital structure. Also, from Eq (15.44), we know that if the common stock ofa firm is thought of as a call option on the value ofthe firm, then as Boy wy 7 Ne). ‘These two facts imply that B ; Sy a N(-ah) = 1 Nid. (15.50) In other words, any change inthe value of equity is offset by an equal and opposite change in the value of risky debt. Next, the required rate of return on risky debt, ky, cam be expressed by using the CAPM, Bq. (15.37); ky = Ry + (Rn — Ry IB ssl) See Gala and Masulis (1976, footnote 15) 586 ‘Chapter 15: Capital Structure and the Cost of Capital: Theory and Evidence Substituting Eqs, (15.49) and (15.50) into (15.51), we have v Ky = Ry + [Ry — RABUN (di) From the CAPM, we know that Ry ~ Ry = (Ry ~ Ry By. ‘Therefore 7 y= Ry + (Ry — R)N-d)e And since Ry =p, v B=Re + (PRYING (15.52) Note that Eq. (15.52) expresses the cost of risky debt in terms of the (OPM. The required rate of return on risky debt is equal tothe risk-free rate, Ry, plus a risk premium, d, where wea’. 0= (0 RNA) ‘A numerical example can be used to illustrate how the cost of debt, inthe absence of bankruptcy costs, increases with the firm's utilization of debt. Suppose the current value of a firm, V, is $3 million; the face value of debt is $1.5 million; and the debs will mature in T = 8 years. The variance of returns on the firm's assets, 02, is 09, its required retum on assets isp = .12; and the riskless rate of interest, A, is 5%. From the Black-Scholes option pricing model, we know that In(V/D) +R, = i UD UE Lae ovT 2 5) +0548) ee + 58 3ve o _ 631+ 4 ©8485 From the cumulative normal probability table (Table 7.7), the value of M(—17125) is approx- imately .0434, Substituting into By, (15.33), we see that the cost of debt is inereased from the risk-free rate, 5%, to 5.61%: + AMS = 1.7125, ky = 05 + (12 — 05)(.0434) 7" = .05 +0061 =.0561, Figure 15.8 shows the relationship of the cost of debt and the ratio of the face value of debt to the current market value of the firm. For low levels of debt, bankruptey risk is trivia, and therefore the cost of debt is close to the riskless rate. tt rises as D/V increases until & equals 6.3%, when the face value of debt, due eight years from now, equals the current market value of the firm. D. The Cost of Capital with Risky Debt 587 Figure 15.8 The cost of % risky debt st ce 06 & 05: T 4 ‘To arrive at a weighted average cost of capital, we multiply Eg. (15.52), the cost of debt, by the percentage of debt in the capital structure. B/V. then add this result to the cost of equity, Eg, (15.48) multiplied by $/V. the percentage of equity in the capital structure. The result is v gt delay tip RIM ar +] ay 4 NG ~ Rd vl [n a =h( *) +p — RAIN(-d) + CAI) =Rytlo- Rp) [l= Ma) + Nd] = 11353) Equation (15.53) is exactly the some as the Modighiani-Miller proposition that in a weskd without tanes the weighted average cost of capital is invariant io changes inthe capital structure ofthe ie Also, simply by rearranging terms, we have + (phy) (1554) 2 $ This is exactly the sare as Eg (15.18), he Modigliani Miller definition ofthe cost of equity capital in a world without taxes. Therefore if we assume that debt is risky, then the OPM, the CAPM, and the Modigliani-Mitler definition are all consistent with one anther. This result is shown sraphically in Fig. 15.9(a). This figure is very similar to Fig, 15.2, which showed the cost of capital ass function ofthe debt to equity ratio, B/S, assuming rskless debs. The only differences between the two figures are that Fig, 15.9 has the debt to value ratio, B/(B + $), ‘on the horizontal axis and it assumes risky debt. Note that in Fig. 15.9(a) the cost of debt increases as more debtis used in the firm's capital structure. Also, ifthe fim were to become 100% debt (not realistic altemative), dhen the cost of debt would equal the cost of capital for an all-equity ir, p. Figure 15.9(b) depicts the weighted average cost of capital in a world with corporate taxes only. ‘The usual Modigliani-Miller result is shown, namely, that the weighted average cost of capital declines monotonically as more debt is employed in the capital structure of the firm. The fact that debt is risky does not change any of our previous results. 588 Chapter 15: Capital Structure andthe Cost of Capital: Theory and Evidence Figure 15.9 The cost of capital given risky deb: (a) ro taxes; (b) only corporate taxes. % % B WACC= pll~ 6, 54 5)! wea’. Ky=B)+(p- RNA) : SS a hae ie a) o 3. The Separability of Investment and Financing Decisions ‘A fundamental assumption of the Modigliani-Miller approach (o the capital structure is that the operating cash flows are unaffected or independent of the choice of capital structure. In the last decade or so, this assumption has come into question because the answer changes if itis relaxed. ‘An example, perhaps, was the demise of Allied Federated Department Stores. As they became overburdened with debt, their suppliers began to refuse to extend trade credit. Consequently. the shelves became bare and customers stopped shopping there. This exemple shows that the financial structure of the firm clearly affects its revenues. These effects have come 10 be called business disruption costs. They include a wide range of so-called masket imperfections ranging from reduced sales to investment opportunities that are foregone. ‘Next, suppose that projects carry with them the ability to change the optimal capital structure of the firm as a whole.'” Suppose that some projects have more debt capacity than others (perhaps because they are more flexible due to the real options that are imbedded in them). Then the investment and financing decisions cannot be “handled” as if they were independent. There is very little in the accepted theory of finance that admits of this possibilty, but it cannot be disregarded. One reason that projects may have separate debt capacities is that they have different collateral values in bankruptcy, or differences in their ability to respond to new information such 2s unanticipated demand. A Mode with Business Disruption and Tax-Deductible Interest Leland [1994] and Leland and Toft [1996] have modeled the value of 2 firm assuming thatthe present value of business disruption costs and the present value of lost interest tax shields are "This may be paricubvly elevan when 2 fim is considering a conglomerate merger with anther im in completely diferent ine of busines mith a completely diferent optimal capital structure. E. AModel with Business Disruption and Tax-Deductible Interest. 589 affected by the firm's choice of capital structure. The result is an optimal capital structure that is defined by 2 trade-off between the value created by the present value ofthe interest tax shield, and the value lost from the present value of business disruption costs as well as the present value of Jost interest tax shields. Leland’s work begins with the assurmption thatthe value ofthe unlevered firm, V, follows a diffusion process with a rate of return Sau, thdt +o0W. (1335) Any claim that pays a nonnegative coupon, C,, when the finn is solvent, with value F{V, £) must satisfy the partial differential equation JV RV. HVE =r WiHtRV +0 (15.56) When tis security has no explicit time dependence, then F, (V, equation simplifies to be 0, and the partial differential dev A yy(V) + eV Fy(V) =rFV)+C=0 (1359) with a general sotion F(V) = Ag+ AV + AgV 202? (15.58) To make Eq, (15.58) more concrete, we can apply it to various securities by specifying the appropriate boundary conditions. Let us stat by applying it othe fim’s deb. To do so, we define V,, a the level of asset value at which business disruption occurs and a as the Faction ofthe value of the firm lost to business disruption costs, leaving debt holders with (j — a) Vp, and leaving shareholders with nothing. The boundary conditions are!S MVs0p BV)=(1-a¥p (15.598) AVw BV)=C/r. (15.596) Using the second boundary condition and applying it o the generic valuation Eg, (15.58) we see that Ay =O and Ag = C/r, and we can rewrite (15.58) as it applies specifically to risky debt: BV) = Ag + A\W + AyV (15.60) Next, we observe that asthe value of assets approzches infinity al the second boundary condition, V 00, the value of debt, D(V), approaches the present value of its perpetual fixed coupon stream: B(V)=Ay=C/r ifandonlyif 4; =O and tim V-@"™ =0, Note tat business disruption cosis are assumed to be proportional vo the asset value where business disruption occurs, ‘Thus, if Vy = 0, then business disruption costs are aso 2er0, 590 ‘Chapter 1g: Capital Structure andthe Cost of Capital: Theory and Evidence At the first boundary condition, we know that B(V) = (1 — a) Vp; therefore we cap rewrite (15.60) as follows: BV) = Ag + Av-@? since Ay=0 SC jr AgV"@N8 = (1ma)Vy since Ay =C/r We can now solve for Ay: 1a) — Cpe", and since V = Vp at the boundary, Eg. (15.60) reduces to BU) = C/r + [(1—e)Vp — Cr (VW) 2 = (= par + pall) Vp) (15.61) pp (W/V_y 2", and can be interpreted as the present value of $1 contingent on future business disruption. We can interpret Bq, (15.61) as the present value of risky debt, with two parts, namely, the present value of rskless debt weighted by (one minus) a business disruption facto that reflects both the cost of disruption an its timing, plus the payout if business dstuption occurs also weighted by the same business disruption factor. Next, we examine the effect of the debt tax shield and of expected! business disruption costs. First, consider business disruption a8 a “security” that pays no coupon, bat has a value equal 10 business disruption costs, Vp. at V = Vg. Its value, DC(V), must also satisfy the conditions of Eq, (15.58), bu with different boundary conditions AtV=V, DCW) =e¥y (15.624) ALY 00 DEW) > 0, (15.626) AAS before, we start with Eq, (15.58) and interpret it given the boundary conditions for DC(V), as follows: DCW) = Ay tA + AV Note hatas the value of the assets approaches infinity, then the present value of business disruption costs, DC(V), approaches zero if and only if in the above equation both Ag and A, are equal to ‘zero. At the first boundary condition V = Vp; therefore 4,V-2?? = ap, E, A Medel with Business Disruption and Tax-Deductible interest $e and therefore, at V Aye (f nl ‘The interpretation of this results thatthe present value of expected business disruption costs. their ‘magnitude if business disruption oceurs, « Vg, multiplied by the present value of $1 conditional ‘on future business disruption, so that the present value of expected business disruption can be writen as DC(V) = aVp(¥/¥p) (15.63) Finally, we must consider the present valve of the interest tax shield as 2 “security” that pays ‘4 constant coupon equal to the tax-shellering value of interest payments (TC) as long as the firm is solvent, Its value TB(V) also must satisfy Eq, (15.58), but with the following boundary conditions:!? AVS Vy, TRV) =0 (19 64a) AtV-+00 TRV) =T.C/r), (15.640) Revwriting Eq, (15.58) for the present value of the interest tax shield on debt, we have TBV)=Ag + AV +4 (15.65) Note that as V approaches infinity, the value of the tax shield benefit approaches the tax rate times the present value of debt, AsV co then TB(V) > 7,(C/r) if and only if Ay = T,(C/r) and Ay = 0. Furthermore, at V = Vg we have TBIV)=0= TAC r) ~ (TAC/OUV/V—) (15.66) Putting this all together, we have the conclusion that the total value of the firm has three parts, First is the firm's asset vaive (ie., the volue of the firm ifit had no debt, the unlevered firm, Vy(V)) To this we add the value of the tax benefit from the deductibility of interest payments, T.B(V ), and subtract the present value of business disruption costs, DC(V). Mathematically, this can be written as viv) Vy(V) +7, BV) ~ DEW) = Vol) + TAC HPL = (V/V) 22 | aVpl Vg 20 = Vy(V) +7,B ~ pgl,B-aV ppp (1567) | Neto that Leland [1994] assumes thatthe firm receives the full tax shelter benefit as long as iti solvent. Actually the ‘boundary should be EBIT > C. He handles this case later in his paper. However, since noninterest tax shields are also affected by nt operating loss carry forwards, the model becomes even more complicate. 592 ‘Chapter 15: Capital Structure and the Cost of Capital: Theory and Evidence Figure 15.10 Optimal capital suucture as a trade-off etwcen the interest tax shield and business disruption costs, Present value (dolass) Financial distress costs ett Optimal debt ratio Capital Equation (15.67) has four terms, The first wo are the same as the Modigliani-Miller vatue of levered firm in a world with corporate axes. The third is the expected present value of interest tax shields losts the firm decides to carry more debt and less equity. The fourth is the expected present value of the busines disruption costs that are incurred asthe fim takes on a greater percentage of Di, the market perceives the firm to be successfu, and vice versa For the signaling equilibrium to be established, (1) the signals must be utambiguous (1... when investors observe D > Ds, the firm is always type A), and (2) managers must have incentive to always give the appropriate signal. Ifthe end-of-period value of a successful type-A firm is Vi, and is always greater than the Value of an unsuccessful type-B firm, Vip, then the compensation of the management of atype-A firm is (1569) yt th +yVig iD < De (lie). ate es FATE AWM De

You might also like